Wbiplot: Weighted Biplot

In analytics: Regression Outlier Detection, Stationary Bootstrap, Testing Weak Stationarity, NA Imputation, and Other Tools for Data Analysis

Description

Wbiplot produces a biplot with any weight distribution between Row and Column markers. This way the full spectrum from perfect row resolution (Row-metric preserving biplot) to perfect column resolution (Column-metric preserving biplot) is available.

Usage

Wbiplot(df, numer1, denom1 = 1, numer2, denom2 = 1, cx = 0.5)

Arguments

df	a dataframe with numeric values only
numer1	numerator of first exponent (can be a decimal)
denom1	denominator of first exponent (default: 1)
numer2	numerator of second exponent (can be a decimal)
denom2	denominator of second exponent (default: 1)
cx	graphical magnification factor (default: 0.5)

Details

This function makes use of function Matpow from package powerplus to be able to raise any valid matrix (see Matpow documentation) to any real power between 0 and 1 included.

Value

A biplot of a dataframe with the specified weights. Weights can either be supplied as two fractions, or as two decimal numbers.

Author(s)

Albert Dorador

See Also

Matpow

Examples

require(graphics)

# Exemple 1: Row metric preserving
Wbiplot(USArrests, numer1 = 1, numer2 = 0, cx = 0.6)

# Exemple 2: Column metric preserving
Wbiplot(USArrests, numer1 = 0, numer2 = 1, cx = 0.6)

# Comparison with function \code{biplot} from package \pkg{stats}
biplot(princomp(USArrests), cex = 0.6)

# Example 3: Custom, 50-50
Wbiplot(USArrests, numer1 = 0.5, numer2 = 0.5)

# Example 4: Custom, 20-80
Wbiplot(USArrests, numer1 = 0.2, numer2 = 0.8)

analytics documentation built on May 2, 2019, 3:37 p.m.